Chunk #23 — Methods — Empirical significance thresholds (H0 simulations) and power

Source

A comparison of multivariate genome-wide association methods.

Embedded

yes

Text

The power is defined as the percentage of 1,000 replicates for which the extracted p-value was smaller or log10 BF was larger than the empirical significance thresholds, ensuring an equal type I error rate of 5% for all methods.